[Tests] Re-enable previously skipped e2e/example tests (#5641)

Signed-off-by: wangyu <410167048@qq.com>
Signed-off-by: [Your Name] <your.email@example.com>
Signed-off-by: wangyu <53896905+yenuo26@users.noreply.github.com>

.buildkite/npu/test-npu-ready.yml

@@ -43,7 +43,7 @@ steps:
     depends_on: upload-ready-pipeline
     steps:
       - label: ":sound: TTS · Qwen3-TTS CustomVoice Test"
-        mirror_hardwares: a2b3_npu_1
+        mirror_hardwares: a3_npu_2
         env:
           VLLM_WORKER_MULTIPROC_METHOD: spawn
           VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"

tests/e2e/offline_inference/test_ming_tts_expansion.py

@@ -35,7 +35,6 @@ MIN_AUDIO_SAMPLES = 1000
 pytestmark = [
     pytest.mark.slow,
     pytest.mark.tts,
-    pytest.mark.skip(reason="https://github.com/vllm-project/vllm-omni/issues/4704"),
 ]
 
 

tests/e2e/offline_inference/test_moss_tts_v1_5_expansion.py

@@ -28,10 +28,14 @@ _OMNI_RUNNER_PARAM = (
     {"stage_init_timeout": 600, "trust_remote_code": True},
 )
 
+_SKIP_ISSUE_6417 = pytest.mark.skip(
+    reason="https://github.com/vllm-project/vllm-omni/issues/6417",
+)
+
 pytestmark = [
-    pytest.mark.skip(reason="https://github.com/vllm-project/vllm-omni/issues/4643"),
     pytest.mark.slow,
     pytest.mark.tts,
+    _SKIP_ISSUE_6417,
     pytest.mark.parametrize("omni_runner", [_OMNI_RUNNER_PARAM], indirect=True),
 ]
 

tests/e2e/offline_inference/test_qwen2_5_omni_autoround_w4a16_expansion.py

@@ -36,8 +36,8 @@ BASELINE_MODEL = os.environ.get("QWEN2_5_OMNI_BASELINE_MODEL", BASELINE_MODEL)
 
 _CI_DEPLOY = get_deploy_config_path("ci/qwen2_5_omni.yaml")
 
-_SKIP_ISSUE_4663 = pytest.mark.skip(
-    reason="https://github.com/vllm-project/vllm-omni/issues/4663",
+_SKIP_ISSUE_5652 = pytest.mark.skip(
+    reason="https://github.com/vllm-project/vllm-omni/issues/5652",
 )
 
 
@@ -65,7 +65,7 @@ quant_params = [(QUANTIZED_MODEL, stage_config)]
 # ------------------------------------------------------------------
 
 
-@_SKIP_ISSUE_4663
+@_SKIP_ISSUE_5652
 @hardware_test(res={"cuda": "L4"}, num_cards=4)
 @pytest.mark.parametrize("omni_runner", quant_params, indirect=True)
 def test_text_to_text(omni_runner, omni_runner_handler):
@@ -84,7 +84,7 @@ def test_text_to_text(omni_runner, omni_runner_handler):
 # ------------------------------------------------------------------
 
 
-@_SKIP_ISSUE_4663
+@_SKIP_ISSUE_5652
 @hardware_test(res={"cuda": "L4"}, num_cards=4)
 @pytest.mark.parametrize("omni_runner", quant_params, indirect=True)
 def test_audio_to_text(omni_runner, omni_runner_handler):
@@ -108,7 +108,7 @@ def test_audio_to_text(omni_runner, omni_runner_handler):
 # ------------------------------------------------------------------
 
 
-@_SKIP_ISSUE_4663
+@_SKIP_ISSUE_5652
 @hardware_test(res={"cuda": "L4"}, num_cards=4)
 @pytest.mark.parametrize("omni_runner", quant_params, indirect=True)
 def test_image_to_text(omni_runner, omni_runner_handler):
@@ -130,7 +130,7 @@ def test_image_to_text(omni_runner, omni_runner_handler):
 # ------------------------------------------------------------------
 
 
-@_SKIP_ISSUE_4663
+@_SKIP_ISSUE_5652
 @hardware_test(res={"cuda": "L4"}, num_cards=4)
 @pytest.mark.parametrize("omni_runner", quant_params, indirect=True)
 def test_video_to_text(omni_runner, omni_runner_handler):
@@ -152,7 +152,7 @@ def test_video_to_text(omni_runner, omni_runner_handler):
 # ------------------------------------------------------------------
 
 
-@_SKIP_ISSUE_4663
+@_SKIP_ISSUE_5652
 @hardware_test(res={"cuda": "L4"}, num_cards=4)
 @pytest.mark.parametrize("omni_runner", quant_params, indirect=True)
 def test_mix_to_audio(omni_runner, omni_runner_handler):

tests/e2e/offline_inference/test_qwen3_omni_autoround_w4a16_expansion.py

@@ -27,8 +27,8 @@ pytestmark = [
     pytest.mark.omni,
 ]
 
-_SKIP_ISSUE_3195 = pytest.mark.skip(
-    reason="https://github.com/vllm-project/vllm-omni/issues/3195",
+_SKIP_ISSUE_5652 = pytest.mark.skip(
+    reason="https://github.com/vllm-project/vllm-omni/issues/5652",
 )
 
 QUANTIZED_MODEL = "Intel/Qwen3-Omni-30B-A3B-Instruct-int4-AutoRound"
@@ -77,7 +77,7 @@ quant_params = [(QUANTIZED_MODEL, stage_config)]
 # ------------------------------------------------------------------
 
 
-@_SKIP_ISSUE_3195
+@_SKIP_ISSUE_5652
 @hardware_test(res={"cuda": "H100"}, num_cards=2)
 @pytest.mark.parametrize("omni_runner", quant_params, indirect=True)
 def test_text_to_text(omni_runner, omni_runner_handler):
@@ -96,7 +96,7 @@ def test_text_to_text(omni_runner, omni_runner_handler):
 # ------------------------------------------------------------------
 
 
-@_SKIP_ISSUE_3195
+@_SKIP_ISSUE_5652
 @hardware_test(res={"cuda": "H100"}, num_cards=2)
 @pytest.mark.parametrize("omni_runner", quant_params, indirect=True)
 def test_audio_to_text(omni_runner, omni_runner_handler):
@@ -120,7 +120,7 @@ def test_audio_to_text(omni_runner, omni_runner_handler):
 # ------------------------------------------------------------------
 
 
-@_SKIP_ISSUE_3195
+@_SKIP_ISSUE_5652
 @hardware_test(res={"cuda": "H100"}, num_cards=2)
 @pytest.mark.parametrize("omni_runner", quant_params, indirect=True)
 def test_image_to_text(omni_runner, omni_runner_handler):
@@ -142,7 +142,7 @@ def test_image_to_text(omni_runner, omni_runner_handler):
 # ------------------------------------------------------------------
 
 
-@_SKIP_ISSUE_3195
+@_SKIP_ISSUE_5652
 @hardware_test(res={"cuda": "H100"}, num_cards=2)
 @pytest.mark.parametrize("omni_runner", quant_params, indirect=True)
 def test_video_to_text(omni_runner, omni_runner_handler):
@@ -164,7 +164,7 @@ def test_video_to_text(omni_runner, omni_runner_handler):
 # ------------------------------------------------------------------
 
 
-@_SKIP_ISSUE_3195
+@_SKIP_ISSUE_5652
 @hardware_test(res={"cuda": "H100"}, num_cards=2)
 @pytest.mark.parametrize("omni_runner", quant_params, indirect=True)
 def test_video_to_audio(omni_runner, omni_runner_handler):
@@ -185,7 +185,7 @@ def test_video_to_audio(omni_runner, omni_runner_handler):
 # ------------------------------------------------------------------
 
 
-@_SKIP_ISSUE_3195
+@_SKIP_ISSUE_5652
 @hardware_test(res={"cuda": "H100"}, num_cards=2)
 @pytest.mark.parametrize("omni_runner", quant_params, indirect=True)
 def test_mix_to_audio(omni_runner, omni_runner_handler):

tests/e2e/online_serving/test_cosyvoice3_tts_expansion.py

@@ -18,7 +18,11 @@ from tests.helpers.media import load_test_audio_data_url
 from tests.helpers.runtime import OmniServerParams
 from tests.helpers.stage_config import get_deploy_config_path
 
-pytestmark = [pytest.mark.slow, pytest.mark.tts]
+_SKIP_ISSUE_6416 = pytest.mark.skip(
+    reason="https://github.com/vllm-project/vllm-omni/issues/6416",
+)
+
+pytestmark = [pytest.mark.slow, pytest.mark.tts, _SKIP_ISSUE_6416]
 
 MODEL = "FunAudioLLM/Fun-CosyVoice3-0.5B-2512"
 
@@ -66,7 +70,6 @@ tts_async_chunk_server_params = [
 
 
 @hardware_test(res={"cuda": "H100"}, num_cards=1)
-@pytest.mark.skip(reason="https://github.com/vllm-project/vllm-omni/issues/4644")
 @pytest.mark.parametrize("omni_server", tts_server_params, indirect=True)
 def test_voice_clone_zh_001(omni_server, openai_client) -> None:
     """
@@ -89,7 +92,6 @@ def test_voice_clone_zh_001(omni_server, openai_client) -> None:
 
 
 @hardware_test(res={"cuda": "H100"}, num_cards=1)
-@pytest.mark.skip(reason="https://github.com/vllm-project/vllm-omni/issues/4644")
 @pytest.mark.parametrize("omni_server", tts_async_chunk_server_params, indirect=True)
 def test_voice_clone_zh_002(omni_server, openai_client) -> None:
     """
@@ -113,7 +115,6 @@ def test_voice_clone_zh_002(omni_server, openai_client) -> None:
 
 
 @hardware_test(res={"cuda": "H100"}, num_cards=1)
-@pytest.mark.skip(reason="https://github.com/vllm-project/vllm-omni/issues/4644")
 @pytest.mark.parametrize("omni_server", tts_server_params, indirect=True)
 def test_voice_clone_en_001(omni_server, openai_client) -> None:
     """

tests/e2e/online_serving/test_higgs_audio_v3.py

@@ -48,6 +48,10 @@ DEFAULT_SPEECH_TIMEOUT_S = 180.0
 # silence-only outputs without flagging short legitimate clips.
 _MIN_AUDIO_BYTES = 20_000
 
+_SKIP_ISSUE_6418 = pytest.mark.skip(
+    reason="https://github.com/vllm-project/vllm-omni/issues/6418",
+)
+
 # Reuse the shared TTS reference clip (clean ~5 s 24 kHz mono human speech)
 # vendored under tests/assets/qwen3_tts/. Keeps a single WAV across TTS
 # tests rather than duplicating asset bytes.
@@ -78,6 +82,7 @@ class TestHiggsAudioV3OnlineHappyPath:
             }
         )
 
+    @_SKIP_ISSUE_6418
     @pytest.mark.slow
     @pytest.mark.tts
     @hardware_test(res={"cuda": "L4"}, num_cards=1)
@@ -95,7 +100,7 @@ class TestHiggsAudioV3OnlineHappyPath:
             }
         )
 
-    @pytest.mark.skip(reason="issue#4411")
+    @_SKIP_ISSUE_6418
     @pytest.mark.slow
     @pytest.mark.tts
     @hardware_test(res={"cuda": "L4"}, num_cards=1)
@@ -112,7 +117,6 @@ class TestHiggsAudioV3OnlineHappyPath:
             request_num=3,
         )
 
-    @pytest.mark.skip(reason="issue#4411")
     @pytest.mark.slow
     @pytest.mark.tts
     @hardware_test(res={"cuda": "L4"}, num_cards=1)
@@ -144,7 +148,6 @@ class TestHiggsAudioV3OnlineHappyPath:
             }
         )
 
-    @pytest.mark.skip(reason="issue#4411")
     @pytest.mark.slow
     @pytest.mark.tts
     @hardware_test(res={"cuda": "L4"}, num_cards=1)
@@ -187,6 +190,7 @@ class TestHiggsAudioV3OnlineInlineControlTokens:
     They do not assert audio quality (out of scope for CI).
     """
 
+    @_SKIP_ISSUE_6418
     @pytest.mark.slow
     @pytest.mark.tts
     @hardware_test(res={"cuda": "L4"}, num_cards=1)
@@ -203,6 +207,7 @@ class TestHiggsAudioV3OnlineInlineControlTokens:
             }
         )
 
+    @_SKIP_ISSUE_6418
     @pytest.mark.slow
     @pytest.mark.tts
     @hardware_test(res={"cuda": "L4"}, num_cards=1)
@@ -219,6 +224,7 @@ class TestHiggsAudioV3OnlineInlineControlTokens:
             }
         )
 
+    @_SKIP_ISSUE_6418
     @pytest.mark.slow
     @pytest.mark.tts
     @hardware_test(res={"cuda": "L4"}, num_cards=1)
@@ -237,6 +243,7 @@ class TestHiggsAudioV3OnlineInlineControlTokens:
             }
         )
 
+    @_SKIP_ISSUE_6418
     @pytest.mark.slow
     @pytest.mark.tts
     @hardware_test(res={"cuda": "L4"}, num_cards=1)
@@ -253,6 +260,7 @@ class TestHiggsAudioV3OnlineInlineControlTokens:
             }
         )
 
+    @_SKIP_ISSUE_6418
     @pytest.mark.slow
     @pytest.mark.tts
     @hardware_test(res={"cuda": "L4"}, num_cards=1)

tests/e2e/online_serving/test_ming_tts_expansion.py

@@ -18,7 +18,6 @@ from tests.helpers.stage_config import get_deploy_config_path
 pytestmark = [
     pytest.mark.slow,
     pytest.mark.tts,
-    pytest.mark.skip(reason="https://github.com/vllm-project/vllm-omni/issues/4704"),
 ]
 
 MODEL = "inclusionAI/Ming-omni-tts-0.5B"

tests/e2e/online_serving/test_moss_tts_expansion.py

@@ -30,10 +30,14 @@ from tests.helpers.mark import hardware_test
 from tests.helpers.runtime import OmniServerParams
 from tests.helpers.stage_config import get_deploy_config_path
 
+_SKIP_ISSUE_6417 = pytest.mark.skip(
+    reason="https://github.com/vllm-project/vllm-omni/issues/6417",
+)
+
 pytestmark = [
     pytest.mark.slow,
     pytest.mark.tts,
-    pytest.mark.skip(reason="https://github.com/vllm-project/vllm-omni/issues/4700"),
+    _SKIP_ISSUE_6417,
 ]
 
 MODEL = "OpenMOSS-Team/MOSS-TTS-Realtime"

tests/e2e/online_serving/test_qwen3_tts_customvoice.py

@@ -20,7 +20,6 @@ from tests.helpers.stage_config import (
     get_deploy_config_stage,
     modify_stage_config,
 )
-from vllm_omni.platforms import current_omni_platform
 
 MODEL = "Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice"
 
@@ -92,10 +91,6 @@ def test_default_cuda_graph_startup(omni_server) -> None:
 @pytest.mark.core_model
 @pytest.mark.advanced_model
 @pytest.mark.tts
-@pytest.mark.skipif(
-    current_omni_platform.is_npu(),
-    reason="#issue 5479",
-)
 @hardware_test(res={"cuda": "L4", "npu": "A3"}, num_cards=1)
 @pytest.mark.parametrize("omni_server", tts_server_params, indirect=True)
 def test_text_to_audio_001(omni_server, openai_client) -> None:
@@ -122,10 +117,6 @@ def test_text_to_audio_001(omni_server, openai_client) -> None:
 @pytest.mark.core_model
 @pytest.mark.advanced_model
 @pytest.mark.tts
-@pytest.mark.skipif(
-    current_omni_platform.is_npu(),
-    reason="#issue 5479",
-)
 @hardware_test(res={"cuda": "L4", "npu": "A3"}, num_cards=1)
 @pytest.mark.parametrize("omni_server", tts_server_params, indirect=True)
 def test_text_to_audio_002(omni_server, openai_client) -> None:

tests/e2e/online_serving/test_sd3_expansion.py

@@ -51,10 +51,7 @@ def _get_diffusion_feature_cases(model: str):
                     "2",
                 ],
             ),
-            marks=[
-                *FOUR_CARD_FEATURE_MARKS,
-                pytest.mark.skip(reason="#3432"),
-            ],
+            marks=FOUR_CARD_FEATURE_MARKS,
         ),
     ]
 

tests/examples/offline_inference/test_text_to_image.py

@@ -26,9 +26,6 @@ EXAMPLE_OUTPUT_SUBFOLDER = "example_offline_t2i"
 def _skip_readme_snippet(language: str, code: str, h2_title: str) -> tuple[bool, str]:
     if h2_title == "Web UI Demo":
         return True, f"README section '{h2_title}' is intentionally excluded for examples tests"
-    # Krea-2 Turbo/Raw README snippets currently fail in CI (more_cli_examples_006/007).
-    if "krea/Krea-2-Turbo" in code or "krea/Krea-2-Raw" in code:
-        return True, "Krea-2 README snippets are temporarily skipped (CI failures)"
     return False, ""